MS Vienna 2398...

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi everyone,

I've spent the day in the British Library, closely examining 15th Century 
substitution ciphers.

The documents I looked at were:-
(1) the BL's facsimile of MS Vienna 2398 (the set of ciphers compiled by 
Francesco Tranchedino for the Sforza chancellery)
(2) Lydia Cerioni "La diplomazia sforzesca nella seconds meta del 
Quattrocento e i suoi cifrari segreti" (1970), which is in two volumes, the 
second volume comprising not only MS Vienna 2398, but also "di cifrari 
conservati presso l'Archivio di Stato di Milano (tavv. I - LXXII)"
(3) a number of other VMS-related documents (but more on those another day).

For those who wish to look at these themselves, the quality of the 
facsimile in (1) is massively better than (2): but (2) makes up for it by 
not only including a ton of additional ciphers from Milan from the same 
period, but also meticulously cross-referencing (and giving biographies of) 
all the people involved.

The structure of MS Vienna 2398 is quite different from what I was lead to 
believe: rather than being a compendium of codes assembled by a single 
person (Tranchedino), it is a document that grew by accreting layers over 
50 years. Most ciphers are dated and named (ie, where they came from, or 
who they were designed for), and their dates range from 1450 to 1496 
(Tranchedino died in Milan in 1481, and the document's handwriting changes 
in 1476: so it's not completely improbable that he fell ill then).

Also: there are hardly any astrological symbols in the substitution 
alphabets, and there are definitely no clusters (or sequences) of 
astrological substitution alphabets. This is true of both MS Vienna 2398 
and the Milanese ciphers. And there are also absolutely no gallows 
characters to be seen anywhere either.

Structurally, most of the ciphers have nulls (marked "nulle", or "nihil 
importantes"), as well as special codes for doubled letters ("duplicate", 
duple", or "gemine"), and alternatives for frequent letters (vowels almost 
always), as well as special symbols for cribs, like "perche", "Papa", or 
"King of France" (etc).

Therefore, even if the VMS were dated to 1450, we should not be surprised 
if doubled letters had been suppressed in some way. But, even so, there 
doesn't appear to be any major similarities with any of the alphabets used: 
so that would appear to be the end of this avenue.

But actually, there is one thing I noticed that I found rather subtle and 
strange. In three very early Tranchedino ciphers, the joined-up character 
"4o" appears. True, you'd expect "40" to be found in number-based ciphers 
or codes... but these aren't really numerical ciphers.

The cipher on page f3v (dated 25th October 1450, "cum m.co d. Tristano 
Sfortia" ie, "Sforza") is a simple substitution cipher, replacing plaintext 
"c" with "4", and "s" with "4o". It may just be Tranchedino's handwriting, 
but it does look a lot like the joined-up VMS "4o".

Similarly, the cipher on page f10v (from 1455, "cum Orpheo de Rycavo" - the 
commentary says "il destinatario si trovata forse a Roma") replaces 
plaintext "b" with "4" or "3", and "u" with "4o", "6o" or another.

And the cipher on page f11r (from 1456, "cum Nicolao Maoleone Ferrariensi") 
maps "a" --> "1/2/3/4", and "i" --> "9/2o/4o/3o".

OK, the 1456 cipher is more number-based: but the *only* two-digit number 
(or conjoined character-pair of any sort) in the 1450 cipher is "4o". Which 
really does demand two questions:-

(a) What is so special about "4o", that it should have been used in the 
1450 cipher as well as the VMS?
(b) Did Tranchedino and/or the Sforza chancellery devise the 1450 cipher 
for Tristano Sforza, or did they merely record it, it having been devised 
elsewhere?

Cheers, .....Nick Pelling.....